Severe flooding in Pakistan contributes to hundreds of deaths, destruction of infrastructure, and the displacement of millions of people. In a new study, an international team of researchers used open data and tools to help understand differences in vulnerability to floods and better predict displacement in the future.
In the summer of 2022, Pakistan experienced severe flooding that displaced approximately 7 million people in the Sindh province alone. Nearly a third of the country, particularly in the south, was underwater. This summer, in mid‑August 2025, unusually intense monsoon rains and rare cloudbursts again triggered catastrophic flash floods in Khyber Pakhtunkhwa where more than 300 people died, entire villages were swept away, and further downpours were expected to continue through September.
Led by Mannie Pui Man Kam, a former doctoral student at ETH Zurich under the mentorship of IIASA researcher Lisa Thalheimer and Radley Horton from Columbia University, a group of researchers from ETH Zurich, the University of Lausanne, and Technical University Berlin undertook a study to better understand how floods lead to displacement, with the aim of preparing populations ahead of the next event. The study, published in Earth’s Future, examined both the social and economic effects and how these impacts interact with one another.
“Current methods for assessing flood impacts remain inadequate and fragmented. We argue that combining displacement estimates with other assessments is key to addressing the social consequences of flooding and preparing vulnerable communities before the next disaster,” Thalheimer explains.
Using the open-source software CLIMADA, the team studied how flood characteristics can help predict how many people might be displaced, even when data is scarce. By examining flood depth, the number of people exposed, and how vulnerable they are, the study estimated how many people were forced to move within the country due to the floods. They also looked at how displacement varied across different areas.
The analysis shows that displacement increases sharply beyond a flood depth of 0.67 meters. Interestingly, the researchers also discovered that displacement was less likely in more urbanized areas. The authors conclude that understanding potential displacement alongside other impacts is essential to reducing the consequences of flooding.
“Our research underscores the critical need to understand the multifaceted impacts of flooding on communities. By leveraging impact assessment tools, we can better forecast displacement, advise governments, and prepare vulnerable populations to reduce losses and damages from extreme weather events such as flooding. This information provides crucial insights for urban planners and policymakers alike. This knowledge can also guide more effective flood management and mitigation strategies before the next flood event occurs,” Thalheimer concludes.
Adapted from a press release by ETH Zurich.
Reference
Kam, P. M., Cache, T., Biess, B., Lohrey, S., di Vincenzo, S., McCaughey, J. W., Horton, R. M., and Thalheimer, L. (2025). Advancing human displacement modelling: A case study of the 2022 summer floods in Pakistan. Earth’s Future DOI: 10.1029/2025EF006788
News
10 June 2026
Annual global migration has nearly tripled since 2000
09 June 2026
Life after work: Why social connections matter
02 June 2026